Semantic mapping for articulated objects
Semantic mapping has advanced greatly since its inception as a research field, to now being able to identify poses and segment objects. As an extension of semantic segmentation problem, there is still an unexplored field of identifying joints of an articulated object within an image. In this project...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project (FYP) |
| Language: | English |
| Published: |
Nanyang Technological University
2019
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/136536 |
| _version_ | 1824453728400310272 |
|---|---|
| author | Luar, Shui Song |
| author2 | Justin Dauwels |
| author_facet | Justin Dauwels Luar, Shui Song |
| author_sort | Luar, Shui Song |
| collection | NTU |
| description | Semantic mapping has advanced greatly since its inception as a research field, to now being able to identify poses and segment objects. As an extension of semantic segmentation problem, there is still an unexplored field of identifying joints of an articulated object within an image. In this project, our main contributions are to re-train a semantic segmentation network on a smaller subset of items which can be considered prismatic or revolute. With a DeepLabv3-Inception network with a ResNet101 backbone, we report best pixelwise accuracy of 0.931 and mIOU of 0.606. while training on 2 object classes from the ADE20K dataset. This preliminary result shows the viability of such an approach, and future work might entail exploring different loss functions; different neural network architecture and expanding the definition to encompass more items from the ADE20K dataset. |
| first_indexed | 2025-02-19T03:11:02Z |
| format | Final Year Project (FYP) |
| id | ntu-10356/136536 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2025-02-19T03:11:02Z |
| publishDate | 2019 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1365362023-07-07T16:36:58Z Semantic mapping for articulated objects Luar, Shui Song Justin Dauwels School of Electrical and Electronic Engineering JDAUWELS@ntu.edu.sg Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Engineering::Computer science and engineering::Computing methodologies::Computer graphics Semantic mapping has advanced greatly since its inception as a research field, to now being able to identify poses and segment objects. As an extension of semantic segmentation problem, there is still an unexplored field of identifying joints of an articulated object within an image. In this project, our main contributions are to re-train a semantic segmentation network on a smaller subset of items which can be considered prismatic or revolute. With a DeepLabv3-Inception network with a ResNet101 backbone, we report best pixelwise accuracy of 0.931 and mIOU of 0.606. while training on 2 object classes from the ADE20K dataset. This preliminary result shows the viability of such an approach, and future work might entail exploring different loss functions; different neural network architecture and expanding the definition to encompass more items from the ADE20K dataset. Bachelor of Engineering (Electrical and Electronic Engineering) 2019-12-26T05:26:15Z 2019-12-26T05:26:15Z 2019 Final Year Project (FYP) https://hdl.handle.net/10356/136536 en application/pdf Nanyang Technological University |
| spellingShingle | Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Engineering::Computer science and engineering::Computing methodologies::Computer graphics Luar, Shui Song Semantic mapping for articulated objects |
| title | Semantic mapping for articulated objects |
| title_full | Semantic mapping for articulated objects |
| title_fullStr | Semantic mapping for articulated objects |
| title_full_unstemmed | Semantic mapping for articulated objects |
| title_short | Semantic mapping for articulated objects |
| title_sort | semantic mapping for articulated objects |
| topic | Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Engineering::Computer science and engineering::Computing methodologies::Computer graphics |
| url | https://hdl.handle.net/10356/136536 |
| work_keys_str_mv | AT luarshuisong semanticmappingforarticulatedobjects |